Character and object recognition by using image and sound fusion

In this thesis, a novel approach that integrates image and sound recognition tasks is presented. Human brain is too complex to be analyzed. But some of the main learning tasks in the brain can be simulated. The learning processes in the brain cannot be considered in the absence of sound or image. Both image and sound plays very important role in the learning processes. Therefore in this thesis the new algorithms that integrate the sound and image recognition tasks are tried in order to become closer to the operation of the human brain. The algorithm takes the sound and image input simultaneously, and extracts the meaningful features and then integrates them as a new knowledge in the data field. After the new knowledge is saved then the new image or sound inputs corresponding to the same knowledge are used to teach the algorithm. The user takes role as a teacher in the algorithm. The experimental results show that the proposed approach is applicable on the learning processes.